opus-js-animations
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill includes logic in
scripts/get_audio.pyto download and update theyt-dlputility from its official GitHub releases if it is not already installed on the system. This is a standard and well-known tool used for the skill's primary purpose of retrieving audio for video creation. - [COMMAND_EXECUTION]: The skill makes extensive use of subprocess calls to execute
ffmpeg,ffprobe,yt-dlp, and Google Chrome (headless). These tools are necessary for audio analysis, transcription, and frame-exact video rendering. The commands are constructed using list-based arguments to prevent shell injection vulnerabilities. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data such as audio transcripts and YouTube metadata. To mitigate the risk of malicious instructions in this data influencing the agent, the workflow enforces a "Director's Treatment" phase. This requires the agent to reflect the project brief and obtain explicit user approval before any JavaScript animation code is written or rendered.
- [CREDENTIALS_UNSAFE]: The skill manages API keys for ElevenLabs and OpenAI text-to-speech services. It utilizes secure practices by reading these keys from environment variables or a specific configuration file (
~/.config/opus-js-animations/keys.env) with restricted permissions (chmod 600), and it contains explicit instructions for the agent never to hardcode or disclose these secrets.
Audit Metadata